EP4309264A1

EP4309264A1 - Electric drive system

Info

Publication number: EP4309264A1
Application number: EP23701884.1A
Authority: EP
Inventors: Alexander Rosen; Dean PETROVSKI; Christopher Römmelmayer; Maximilian HABERSBRUNNER
Original assignee: DeepDrive GmbH
Current assignee: DeepDrive GmbH
Priority date: 2022-03-02
Filing date: 2023-01-23
Publication date: 2024-01-24
Also published as: CN117751508A; DE102022202123A1; DE102022202123B4; WO2023165760A1; KR20240033055A

Abstract

The present invention relates to an electric drive system for or in a motor vehicle, having: at least one synchronous machine, which has a double rotor and a distributed winding placed in a stator core, wherein the double rotor is constructed from flux-carrying material made of solid material, wherein the winding is designed to be self-supporting for torque support; and at least one three-stage or multi-stage inverter circuit that is coupled to the synchronous machine at a load output and that is designed to convert a DC voltage received on the supply side into an AC voltage via which the synchronous machine can be driven via the load output, wherein the inverter circuit has a controllable three-stage or multi-stage inverter.

Description

Electric Drive System FIELD OF THE INVENTION The present invention relates to an electric drive system for or in a motor vehicle. TECHNICAL BACKGROUND Electric machines with a stator and two rotors connected to one another in a rotationally fixed manner, so-called double-rotor machines (in addition to double-rotor also referred to as multiple rotor, dual-rotor, etc.) can increase both the torque density and the efficiency of electric drives compared to conventional electric machines increase with only one rotor. This can be explained by the fact that no magnetic yoke is required in the stator, particularly in the so-called “yokeless” design, which means that the core losses can be significantly reduced. In addition, with two rotors there is always more space available for the field-exciting magnets (in the case of permanent-magnet synchronous machines, PSM) or the conductor material (in the case of induction machines, IM or electrically excited synchronous machines, ESM). According to the alignment of the magnetic field lines in the air gap, such machines can be divided into two groups: carrying axial flux (field lines parallel to the axis of rotation, so-called axial flux machines) and carrying radial flux (field lines in the radial direction in the air gap, so-called radial flux machines). ) on the other hand. Axial flow double rotor machines are described for example in DE 102015 226 105 A1 and DE 102013 206 593 A1. They are characterized by a high torque and power density, but are complex to manufacture because very complex geometries are stamped or stamped in the stator core have to be manufactured by powder metallurgy. So far, such machines have not made the leap into mass production and are only used in niche areas with high demands on power density, such as racing, aviation, etc. In addition, the mechanical fastening concepts for the stator winding only allow the use of single-tooth windings with the corresponding disadvantages in terms of noise excitation. In contrast, established manufacturing processes suitable for large-scale production can be used in principle for the winding and laminated core in radial-flow double-rotor machines. However, there is a major and largely unsolved technical challenge in supporting the torque generated in the stator core. Due to the internally and externally rotating parts, the laminated core of the stator cannot be installed in a fixed housing (for example pressed, screwed or glued) as is otherwise usual. The torque is therefore conducted to the axial ends of the stator core or stator winding and supported there. Various approaches are proposed for this purpose in the prior art, but they are all associated with considerable disadvantages in terms of function and/or costs. EP 1879 283 B1 describes a possibility of designing the stator winding as a so-called yoke winding. The ring-shaped laminated core of the stator has grooves on the inner and outer diameters, between which there is a magnetic yoke that is effective in the tangential direction (also referred to as the stator yoke). In this case, the forward and return conductors of each phase winding are guided in slots that lie radially one above the other and are wound around the yoke. The stator yoke is axially accessible between the winding strands and can, for example, be screw connections on the housing (e.g. described in JP 2018082 600). The axial compression of the screws ensures both torsional rigidity of the laminated core and torque support at the axial end. The north and south poles of the rotor field face each other. The disadvantage of this concept is that the magnetic flux has to be routed completely via the return yoke located between the stator slots. On the one hand, this leads to an increased weight of the stator lamination stack and significantly increases the iron losses. The magnetic field lines of both rotor fluxes close via the magnetic return flux in the stator core and cause iron losses there. In addition, all individual coils of the yoke winding in the area of the end winding must be connected in parallel or in series, which in turn leads to a space conflict with the torque support. However, the winding wound around the yoke allows direct mechanical contact with the laminated stator core. A significant reduction in weight and loss can be achieved if the magnetization directions of the magnets lying radially one above the other point in the same direction and the directions of current flow of the conductors lying one above the other in the slots are identical. In this case, the magnetic yoke in the stator can be omitted and a so-called "yokeless" double-rotor machine with distributed winding is created. The magnetic field lines close above the rotor. A magnetic yoke in the stator is not required, which means that the weight and iron losses in such machines are very low. However, the distributed winding does not allow direct mechanical contacting of the laminated core of the stator for torque support. For example, WO 2004/004098 A1 describes a yokeless design with distributed winding. Even with a so-called "yokeless" design, it can still make sense to design a thin yoke for the mechanical connection of the stator teeth, but this is not necessary electromagnetically. The term "yokeless" thus refers to electromagnetic flux conduction in which there is no flux in the tangential direction in the stator. However, the winding cannot be designed as a yoke winding, since the outgoing and return conductors of the winding strands are distributed radially around the circumference and thus form a distributed winding. This results in winding overhangs of distributed windings, which make it difficult to access the laminated core in the axial direction. The purely radial flux guidance also prohibits the use of axial, metallic screw connections, as these form conductor loops with a large number of interlinked fluxes and high additional current heat losses. Various auxiliary constructions for torque support are proposed in the prior art for axial support, for example described in DE 102010 055 030 A1 or US Pat. No. 7,557,486 B2. The problem here is that electrically and/or magnetically conductive metals are not allowed to protrude into the flux-carrying area, or only to a very limited extent, which greatly restricts the choice of material and geometric design. In contrast, plastic components, adhesives and/or potting materials can also be used in the flow-carrying area. With such materials, however, it is very difficult to meet the high requirements in terms of temperature stability and mechanical strength. An inverter is provided for the operation of such a double-rotor machine for or in a motor vehicle due to the DC voltage present at the output of suitable energy storage devices. An inverter, also known as an inverter or rotary converter, is an electrical device that converts direct current into alternating current. Such inverters are used, for example, in modern motor vehicles, in photovoltaics (solar inverters), as components in frequency converters and many other applications in which a suitable AC voltage is to be generated from a DC voltage. Inverters of this type and their areas of application are generally known in a wide variety of circuitry variants, so that there is no need to go into more detail about their circuitry design and mode of operation. In modern motor vehicles, electrically powered drive systems are increasingly being used, also for reasons of sustainability and to avoid CO2 emissions. Such drive systems include, for example, one or more electrical machines, such as synchronous machines or asynchronous machines, which are supplied with a multi-phase AC voltage. So-called two-level inverters (also called 2-level inverters or 2L inverters for short) are generally used to generate the AC voltage. With two-stage inverters, an AC voltage with two voltage levels is generated from the DC voltage of a DC voltage source. Two-stage inverters have prevailed over other inverter topologies, particularly in the area of drive inverters for electric vehicles. Currently, IGBT switching elements are predominantly used in two-stage inverters. An example of such a two-level inverter is given, for example, in the paper by H. v. Hoeck, "Power Electronic Architectures for Electric Vehicle", published in the IEEE in 2010 published book "Emobility - Electrical Power Train". In addition to the two-stage inverter topology just mentioned, there are also three-stage or multi-stage inverter topologies, with which three-stage or multi-stage voltage levels can be generated. Examples of multi-level inverter topologies are described, for example, in US Pat. No. 10,903,758 B2 or US 2017/0185130 A1. The advantages of several voltage levels are lower harmonics, a slower voltage change at the phase outputs, low electromagnetic emissions (EME) and above all the processing of higher voltages. For these reasons, such three- or more-stage inverters are currently used primarily for high-voltage applications. Energy technology applications, such as solar inverters or wind turbines, are established areas of application for such three- or multi-stage inverter topologies. Higher voltages are not to be found in electric vehicles (with voltages of e.g. 400V). In photovoltaics, on the other hand, voltages of more than 1kV are common; with other renewable energies, such as wind energy, the voltages are significantly higher. According to the prevailing opinion, the advantages of three- or multi-stage inverters just mentioned are not sufficient to justify their use in electric drives of electric vehicles, as is the case in the article by Andreas Bubert et. al., "Experimental Validation of Design Concepts for Future EV-Traction Inverters", 2018 IEEE Transportation Electrification Conference and Expo (ITEC), pp. 795-802. For all these reasons, three or more stage inverter topologies are gies in electrically powered vehicles are not used today. SUMMARY OF THE INVENTION The object of the present invention is to improve the efficiency of an electric drive system equipped with a double rotor, while at the same time enabling simpler and more cost-effective manufacture. According to the invention, this object is achieved by an electric drive system having the features of patent claim 1 . Accordingly, there is provided: an electric drive system for or in a motor vehicle, with: at least one synchronous machine, which has a double rotor and a distributed winding placed in a stator core, the double rotor being made of flux-carrying material made of solid material, wherein the winding is designed to be self-supporting for torque support; and at least one three-stage or multi-stage inverter circuit, which is coupled to the synchronous machine at a load output and which is designed to convert a DC voltage taken up on the supply side into an AC voltage, via which the synchronous machine can be driven via the load output, the Inverter circuit has a controllable three- or multi-stage inverter. The core of the present invention consists in a combination of a special electrical synchronous machine with a double rotor, the stator of which is equipped with a distributed winding designed to be self-supporting and whose double rotor is made of solid rotor material, i.e. in full construction, and a three- or multi-stage inverter circuit. For the first time, the torsionally stiff winding offers the possibility of constructing the winding of a synchronous machine with a double rotor, in particular a yokeless double rotor machine, as a distributed winding with a correspondingly low upper field spectrum. Only in this design can the rotors be made of solid or solid material, since the winding only produces small upper fields and the resulting eddy currents in the rotor. The manufacture of rotors is thus massively simplified and is much more cost-efficient. The design of an electric drive system according to the invention thus offers considerable technological and economic advantages because material costs and complexity in the rotor are reduced. However, if such a machine were to be operated with a two-stage (2-level) converter, as is usual with electrical machines in the prior art, the current harmonics induced by the converter would cause eddy currents and additional losses in the rotors made of solid material. These harmonic losses are particularly relevant at low output torques and greatly reduce efficiency. The causative harmonics in the input voltage can be significantly reduced by the inventive use of the combination of rotors made of flux-carrying solid material and a three-stage or multi-stage inverter, which leads to a reduction in losses by more than 75%. Since the harmonic losses in electrical machines in the prior art do not play a major role, de this does not justify the additional costs of a three-stage inverter. The combination of features according to the invention can only be rated as advantageous in this special case. One of the underlying findings here is that electrical machines with a double rotor made of solid material have high losses in the rotor when fed from a conventional 2L inverter. Structurally, the losses in the electrical machine cannot be reduced or can only be reduced to an insignificant extent. Reducing the losses by increasing the frequency in 2L operation has little effect and increases the losses in the inverter, which in turn affects the overall efficiency. The basic mechanism for reducing the losses in the solid material of the double rotor is based on the fact that the amplitude of that magnetic flux density in the solid material of the double rotor which does not contribute to the generation of torque should be reduced. This component, which is defined by harmonics in the flux density, is approximately directly proportional to the square of its amplitude to the change in the THD-induced losses. So changing the inverter switching frequency results in an indirect linear change in the losses and is therefore less effective. A reduction in losses in the solid material contributes significantly to reducing the overall losses of the electrical machine and to its economical use. The resulting knowledge, which is part of the present invention, is that the losses in the electrical machine are compensated by an inverter circuit, which only reduces the amplitude of the harmonics in the flux density can be effectively reduced. In order to achieve this, the following measures and aspects were taken into account when designing and selecting the mode of operation of the inverter: The function of the 2L inverter is replaced by the function of a 3L inverter in order to reduce the harmonics at the phase outputs of the inverter to reduce. This reduces the harmonics in the flux density and in the stator current. A frequency change is not necessary for this. Although the losses are also reduced by increasing the switching frequency in 2L operation, this is not done because the switching losses in the inverter would also increase sharply and the overall efficiency is therefore not significantly improved. Although an increase in the switching frequency could positively support the loss optimization, it is not an essential aspect of the solution according to the invention. The 3L inverter used, on the other hand, offers three voltage levels (3L) and is preferably (but not necessarily) three-phase. With three voltage levels and three phases, a relatively high level of cost efficiency can be achieved. However, the system can be expanded to any number of phases and any number of voltage levels with the same design of all phases. In contrast to known 2L inverters, the power losses in the electrical machine are greatly reduced when operating a 3L inverter according to the invention due to the lower harmonics. The switching losses of the 3L inverter are also reduced comparatively, while the conduction losses are increased. The prevailing loss mechanisms change depending on the load both in the electrical machine and in the 3L inverter. In 3L operation, the harmonics are lower, so the machine losses are greatly reduced. Harmonic induced losses are dominant at low currents. With larger currents, the predominant loss mechanism changes and resistive line or copper losses dominate, whereas losses induced by harmonics are of secondary importance or are comparatively small. Inverter switching losses are reduced (approximately by 50%) in the 3L inverter compared to 2L inverters. With small loads (currents) these switching losses are dominant, whereas with larger currents conduction losses dominate and 2L operation is more efficient. These findings lead to the idea according to the invention of using a 3L inverter for low loads and a 2L inverter for high loads. This operation is possible using the controllable three-stage or multi-stage inverter according to the invention. Overall, the advantages of 2L operation can be combined with the advantages of 3L operation, especially in the case of electric machines equipped with double rotor motors, in order to increase the overall efficiency of the electric drive system compared to known electric drive systems to improve significantly. A further finding on which the present invention is based is that a winding of a radial flux double rotor machine can be designed in a force-transmitting manner for torque support. One of the present The aspect on which the invention is based is therefore to design the winding arranged in the stator core to be self-supporting for torque support. A self-supporting design of the winding means that the winding is sufficiently rigid and strong against torsion around the machine axis to support the drive torque. The self-supporting winding is in particular embedded in a soft-magnetic stator core for magnetic flux guidance. This results in the particular advantage that the stator core itself does not require any inherent torsional rigidity with respect to the machine axis and no other auxiliary construction is required to fix the stator core. Rather, the torque is supported, in particular completely, via the winding. The winding is a so-called distributed winding. This means that the forward and return conductors of the winding strands are distributed around the circumference in the tangential direction in such a way that the conductors of other strands are positioned in the tangential direction between the forward and return conductors of one strand. This results in an interlacing that requires conductors of different strands to cross, particularly in the end winding area. Distributed windings are in particular in contrast to so-called toothed coil windings, in which the outgoing and return conductors of a phase lie in adjacent slots, which leads to winding overhangs without crossing. In this way, a functional integration that was previously unknown or technically unfeasible in the field of radial flux double rotor machines is created, in that the distributed winding is given a supporting function to support the torque in addition to carrying current. sample For this purpose, a mechanical fixation of the winding outside of the stator core can be provided at one axial end. In order to produce such a winding, an integral manufacture of the winding in the existing stator core is proposed. The individual bars of the winding are inserted in the axial direction following the helix of the stator slots through the radially inner and radially outer stator slots and connected at the conductor ends. An integral connection by welding or soldering is preferably provided here. In this way, the winding is positively connected to the stator core. The selected pitch angle (also twisting angle) of the stator slots or the helixes described with it ensures that by connecting the inserted conductor bars, nested conductor loops are formed, which thus form a distributed winding. The angle covered by the conductor loops in the machine in relation to the central axis encloses one magnetic pole of the rotors. In this way, despite the functional integration, a very simple production of the stator is made possible, which manages with very few components and comparatively simple conventional connection technology and therefore also with very few production steps. The stator designed in this way can now be completed with inner and outer rotors made of solid material to form a synchronous machine of an electric drive system according to the invention. For example, these are permanent-magnet-excited rotors with surface magnets and/or buried magnets, squirrel-cage rotors or electrically excited rotors. Hybrid variants with different rotor variants in the inner and outer rotor can also be provided. One A particularly advantageous embodiment results when the rotors are made of soft magnetic solid material and are made with surface-mounted permanent magnets. The low upper field spectrum of the winding variants described here and the distance between the solid material and the air gap guaranteed by the magnets prevent impermissibly large losses from occurring as a result of eddy currents in the rotors. In this embodiment, comparatively high efficiencies can then advantageously be achieved and the rotors can nevertheless be manufactured very inexpensively. The synchronous machine can be integrated into a vehicle axle, for example, and can be provided to drive a drive wheel. In particular, the synchronous machine can be coupled to the drive wheel without a gear. Thus, according to one aspect, an electric drive system according to the invention with a vehicle axle, in particular for a motor vehicle, is also disclosed, wherein the synchronous machine with a double rotor is coupled to a drive wheel in a gearless manner. According to one aspect, a motor vehicle with such an electric drive system is also disclosed. Advantageous refinements and developments result from the further dependent claims and from the description with reference to the figures of the drawing. According to an advantageous embodiment, the synchronous machine is designed as a radial flux double rotor machine. Due to the construction according to the invention, the mass of a radial flux double rotor machine can be reduced and the torque density can be increased due to the functional integration. According to a particularly preferred embodiment, the synchronous machine is designed for a wheel hub drive, in particular as a wheel hub motor for an electrically operated motor vehicle. A wheel hub motor is an electrical machine that is installed directly in a wheel and in particular in the hub of a vehicle and at the same time supports the wheel hub. A part of the hub motor transfers the generated torque directly to the wheel to be driven, with which it rotates. The main advantage of such electric wheel hub motors in vehicles compared to drive concepts with a central motor is the omission of the classic drive train with the necessary components depending on the design (transmission, cardan shaft, differential gear, drive shaft, etc.). Since there are no transmission losses either, there is potential for increasing the efficiency of the entire drive system. Efficient recuperation, ie recovery of electrical energy when braking the vehicle, can also be implemented with an electric wheel hub motor. According to a particularly preferred embodiment, it is a wheel hub drive with a radial flux double rotor machine. The construction according to the invention, which reduces the mass of a radial flux double rotor machine and increases the torque density, enables a particularly advantageous reduction of unsprung masses in a vehicle axle in wheel hub motors. Furthermore, according to the invention, a comparatively short axial length can be realized with a comparatively large diameter, which is advantageous in particular in the wheel interior with regard to torque support and installation space. On the other hand, according to the invention, despite the extremely compact design, very high torques are also possible, which in particular are high enough to directly drive a wheel of a vehicle without a gear. drive. In this way, particularly advantageously, transmission losses are avoided, further weight is saved and particularly high efficiency advantages can be achieved. In addition, this high torque, which is already possible in the four-digit range, in particular greater than 5000 Nm, for sizes within the dimensions of conventional motor vehicle rims, and thus already reaches the area of the adhesion limit of conventional road tires, even a rear axle Replace the wheel brake with the wheel hub motor. Special synergies are thus made possible when used as a wheel hub motor. According to one embodiment, the winding protrudes beyond the stator core at at least one axial end. Furthermore, a support device is provided which is arranged offset axially with respect to the stator core and is designed for positive engagement with the winding at the at least one axial end for torque support. In this way, the self-supporting winding is positively engaged with the carrier device for torque support, which is arranged offset axially with respect to the stator core. According to one embodiment, the synchronous machine has a mechanically fixed base. For torque support, the carrier device is in positive engagement with the at least one axial end of the winding and is supported on the base. The carrier device is firmly connected to the base as the stationary part of the synchronous machine by a suitable method. A possible embodiment provides recesses, for example through-holes, for non-positive fastening means, such as screws. Of course, as an alternative or in addition, form-fitting connecting means and/or a material-to-material connection would also be conceivable. According to one embodiment, the double rotor has a first rotor made of solid material arranged radially inside the stator core and a second rotor made of solid material arranged radially outside of the stator core. The rotors are preferably firmly coupled to one another, for example stamped, riveted or screwed. According to one embodiment, the flux-carrying material in the double rotor or in the first and second rotor consists of iron or an iron alloy. The magnetic flux is thus advantageously optimized. According to a likewise particularly preferred exemplary embodiment, the synchronous machine is a three-phase synchronous machine. In this case, the inverter circuit is preferably designed at least as a three-phase inverter. Another finding of the present invention is that synchronous machines that use a three- or multi-stage inverter topology show a significantly improved overall efficiency of the drive system. According to one embodiment, the inverter circuit has an operating mode setting device which is designed to convert the inverter from three-stage or multi-stage operation to two-stage operation and vice versa, depending on the overall efficiency of the electric drive system. the overall efficiency being a function of the detected phase current of the synchronous machine and at least one other parameter influencing the overall efficiency and/or another property of the synchronous machine influencing the overall efficiency. According to a further embodiment, the inverter circuit has an operating mode setting device which is designed to convert the inverter from three-stage or multi-stage operation to two-stage operation and vice versa, depending on the overall efficiency of the electric drive system the overall efficiency is solely a function of the detected phase current of the synchronous machine or a function of at least one other property of the synchronous machine that influences the overall efficiency. According to one embodiment, the operating mode setting device has an evaluation device which is designed to optimize the overall efficiency on the basis of the phase current or the at least one further property. According to one aspect of the present invention, this has a special inverter circuit connected with an adaptation of the entire drive system, which makes it possible to increase the overall benefit without entailing an increase in costs. To this end, the use of a novel controllable three- or multi-stage inverter is proposed, which can be operated in three- or multi-stage operation (hereinafter referred to as 3L operation) and in two-stage operation (hereinafter referred to as 2L operation). An operating mode setting device provided specifically for this purpose sets the respective operating mode in that the power switches of the inverter are controlled in a suitable manner. The operating mode is set according to the overall efficiency of the entire drive system - and thus not only on the basis of the synchronous machine and/or the inverter used. In addition to the detected phase current of the synchronous machine - as with other inverters - other factors that influence the overall efficiency are also taken into account for the overall efficiency Parameters and/or properties of the synchronous machine are taken into account. The latter is not taken into account in known drive systems for the efficiency consideration and efficiency analysis. According to the invention, an overall efficiency consideration is therefore carried out here. One idea of the present invention consists in reducing the losses, above all in the case of small loads, by the inverter being operated in 3L operation in this case. The losses of the inverter at all operating points are at most insignificantly increased or even reduced. The overall efficiency of the drive system, i.e. the inverter and the synchronous machine, increases significantly, especially when used in electrically powered vehicles. According to a preferred aspect of the present invention, the inverter circuit has an operating mode setting device. What is essential here is that the operating mode setting device does not necessarily switch over sharply from 2L operation to 3L operation and vice versa. Rather, it would also be conceivable if such a switchover instead takes place successively, for example by fading from the inner circuit breakers to the outer circuit breakers taking place. This fading can be carried out, for example, taking into account the average current values of the various circuit breakers, so that the operating times or the times in which the respective circuit breakers are switched on are taken into account. In addition or as an alternative, it would also be conceivable for the circuit breakers to be switched slowly and/or according to a predetermined range sequence. The operating mode setting device, which for example is an evaluation device, a control device and/or measuring has devices can be designed, for example, as a program-controlled device, such as a microprocessor or microcontroller. However, a logic circuit such as an FPGA, PLD or the like would also be conceivable for this function. According to an advantageous development, the operating mode setting device has an evaluation device. The evaluation device is designed to optimize the overall efficiency of the electric drive system using the phase current and using the at least one additional parameter and/or the at least one property of the electric drive system. Typically, but not necessarily, the overall efficiency is calculated numerically by the evaluation circuit. In addition or as an alternative, the overall efficiency can be determined using a predefined family of characteristics, which is mapped in a lookup table, for example. The determination of the overall efficiency can be calculated or determined during operation or in advance. The optimal, i.e. the most efficient operating strategy possible, is preferably calculated in a so-called offline mode before the operation of the electric drive system, for example numerically. This can be accomplished with comparatively little computer resources and is to be preferred above all when a large number of parameters are taken into account in the numerical precalculation of the optimum overall efficiency. In addition, more time is available for the calculation in offline mode. Alternatively, however, a very dynamic determination of the respective operating mode (2L operation or 3L operation) would also be conceivable and possible in a so-called real-time operation, for example via a lookup table. This is This is particularly advantageous and possible if a smaller number of parameters is used to calculate the overall efficiency. For example, one could use a trained artificial network for these purposes, which was trained on the basis of previous parameter values, characteristic curves and the like. According to a preferred refinement, the evaluation device has an optimization module which is designed to initially determine the overall efficiency. Alternatively or additionally, the overall efficiency can then be optimized via an optimization function, taking into account the phase current and the at least one further parameter and/or property. The overall efficiency can be optimized analytically and/or using a suitable lookup table, which has been generated beforehand, for example. At least one of the following parameters is provided as a further parameter: temperature of the inverter circuit; - synchronous machine temperature; - intermediate circuit voltage of the inverter; - rotor speed or rotor speed; - torque of the synchronous machine; - degree of modulation; - Phase voltage or phase current. Other parameters would of course also be conceivable. The operating mode used in each case (e.g. 2L operation or 3L operation) would be, for example, a property of the synchronous machine that influences the overall efficiency. A further property can be seen in the special configuration of the rotor of the synchronous machine, for example such that the rotor is a double rotor and/or that the double rotor is formed from flux-carrying material from solid material. According to a preferred exemplary embodiment, the operating mode setting device has at least one measuring device: A first measuring device has at least one sensor input via which the first measuring device can be coupled to the synchronous machine. The first measuring device is designed to record the phase current, the temperature, the rotor speed and/or other measurable parameters. For example, the temperature of the synchronous machine or its rotors can be recorded using appropriate thermocouples. Alternatively, the change in the temperature-dependent electrical resistance of certain conductors and semiconductors or special semiconductor circuits can be used to generate a voltage that is proportional to the absolute temperature (keyword: bandgap reference). Although the torque of the synchronous machine cannot be measured directly, it can be calculated, among other things, by measuring the phase current. The rotational speed of the rotor and from it the rotor speed can be determined in a variety of ways, for example using a Hall sensor attached to the rotor or an incremental encoder. A second measuring device which is arranged and designed in such a way that it detects the temperature and/or the intermediate circuit voltage of the inverter. The temperature can be recorded analogously to the above with regard to the first measuring device. According to a preferred embodiment, the inverter includes a T-type neutral point clamped (TNPC) inverter architecture. These have various advantages compared to multi-stage Active Neutral Point Clamped (ANPC) inverter topologies: In contrast to ANPC topologies, there are not four but a maximum of three switches conducting in series and the on-state losses are therefore lower. The output voltage forms are identical, which leads to similarly low switching losses, but at higher switching frequencies (e.g. >10kHz) the required total chip area of the TNPC topology is lower compared to the two-level topology. Similar to ANPC, a hybrid inverter topology can also be set up with TNPC in order to further increase efficiency and/or optimize manufacturing costs. For example, different switch technologies can be used for this in the zero-potential or middle bridge branch. Especially in the case of a TNPC inverter built entirely with IGBTs (Insulated Gate Bipolar Transistors), the losses can be drastically reduced using gallium nitride (GaN). The hybrid TNPC inverter topology can also be used in motor controls in electric vehicles, but is not found in practice. It is particularly advantageous to design the 3-level converter as a T-type converter, with the middle switches having a significantly lower current-carrying capacity than the outer switches. In the low output torque range, the converter is then operated in 3L operation and in the high output torque range in 2L operation. The advantage of this design is that the harmonic losses are avoided in the area of low output torques, where they are of particular relevance. TNPC-based 3L inverters can operate in two modes to increase system efficiency. With 3L-TNPC inverters, the zero potential (middle) bridge arms can be switched off to operate in 2L operation and switched on to change to 3L operation. The two operating modes are switched in order to increase the system efficiency. To do this, the load is measured in the control and regulation logic and a switch is made between 2L and 3L operation using a previously determined optimization characteristic. Additionally or alternatively, TNPC-based 3L inverters can be designed asymmetrically to reduce the cost of the inverter. The asymmetry refers to the current-carrying capacity of the zero-potential (middle) bridge branches, which is lower than that of the outer bridge branches. This is possible because the zero-potential bridge arms are no longer used at higher loads in order to optimize overall efficiency. The outer bridge branches are designed for peak currents and the zero-potential bridge branches for small or continuous currents. According to an exemplary embodiment of the invention, the inverter has a first driver stage and at least one second driver stage. The second driver stage is designed to carry output load currents to the load output which are smaller than the output load currents provided by the first driver stage. The operating mode setting device is preferably designed to control the inverter in such a way that, depending on the overall efficiency, the first driver stage and the second driver stage are activated in three-stage or multi-stage operation and at least in two-stage operation. at least one of the driver stages is deactivated, preferably the inner, second driver stage. Typically, but not necessarily, the first driver stage has at least one bridge circuit, in particular a half-bridge circuit, whose center tap forms the output load connection of the inverter circuit. Each bridge circuit has at least one first (semiconductor) power switch, which is connected to a first supply connection (to which a positive supply potential is applied, for example) and which is designed to provide a first voltage stage at the load output. Each bridge circuit also has at least one second (semiconductor) power switch, which is connected to a second supply connection (to which, for example, a negative supply potential or a reference potential is applied) and which are designed to generate a second voltage level at the load output to provide. The semiconductor-based power switches can be implemented with any number of different semiconductor materials. Commonly used materials are Si (silicon) for IGBTs and MOSFETs, SiC (silicon carbide) for MOSFETs and GaN (gallium nitride) for MOSFETs. Typically, but not necessarily, the second driver stage has at least one third power switch, the load paths of which are connected in series between an intermediate circuit circuit and the center tap of the first driver circuit. The power switches of the second driver stage are designed to provide a third voltage level, which lies between the first and the second voltage level, at the load output. In the case of a preferred, so-called homogeneous inverter topology, all the power switches of the inverter, ie the power switches of the first driver stage and/or the second driver stage, are designed as semiconductor switches of the same switch type and/or the same semiconductor technology. Switch types are, for example, bipolar transistors, field effect transistors (such as MOSFETs, JFETs, etc.), thyristors, IGBTs, etc. Semiconductor technology refers to the semiconductor technology on the basis of which the power switch is manufactured, such as on the basis of the Si, SiC, GaAs or GaN technology. In a first, preferred variant of the homogeneous inverter topology, the semiconductor switches are designed as GaN power switches, for example as GaN MOSFETs. In a second, particularly preferred variant, the semiconductor switches are designed as SiC power switches, in particular as SiC MOSFETs. In addition, IGBT-based power switches, for example silicon-based IGBTs with Si diodes or SiC diodes, would also be conceivable. In the case of a particularly preferred, so-called hybrid inverter topology, at least two different switch types and/or at least two different semiconductor technologies are provided for the semiconductor switches of the inverter, i.e. for the semiconductor switches of the first driver stage and/or for the semiconductor switches of the second driver stage. The hybrid inverter topology does not use the same semiconductor materials for all power switches within the inverter. In particular, a different technology (different switch types) is used for the power switches of the zero-potential bridge branch, ie for the second driver stage, than for the outer switches of the first driver stage. As a result, one realizes efficiency advantages due to reduced switching and conduction losses. In addition, there are also cost advantages. In particular, it is recommended to optimize the power switches in the zero-potential bridge branches (second driver stage) for low switching losses and the lowest possible reverse recovery losses. This makes sense since the zero-potential bridge arms (second driver stage) are activated at low currents and low reverse recovery losses also reduce the switching losses in the outer switches. A hybrid design is particularly recommended if the inverter is designed asymmetrically. The lower the current-carrying capacity of the zero-potential bridge branches (second driver stage), the lower the additional costs for switches with optimized switching losses. In a first, particularly preferred variant, the semiconductor switches of the first driver stage are in the form of IGBTs (silicon or SiC) with a freewheeling diode. In this case, the semiconductor switches of the second driver stage can preferably be designed as SiC power switches, in particular as SiC MOSFETs. In a second variant, which is also preferred, the semiconductor switches of the first driver stage are designed as SiC MOSFETs. In this case, the semiconductor switches of the second driver stage can be designed as GaN-based MOSFETs. In a third preferred variant, the semiconductor switches of the first driver stage are designed as IGBTs with a freewheeling diode. In this case, the semiconductor switches of the second driver stage can be designed as GaN power switches, in particular as GaN MOSFETs. According to a particularly preferred embodiment, the flux-carrying material in the rotor consists of iron or an iron alloy. Electrical induction machines - and here preferably synchronous machines with double rotors - can be designed in the rotor with flux-carrying material in solid construction, i.e. made of solid material. The reason for this is that, in an idealized view, there is no periodic relative movement between the directional vector of the rotating field generated by the stator winding and the double rotor in synchronous machines. The magnetic flux density at an operating point is therefore constant and there are no iron losses in the material. In such permanent-magnet machines, whose magnets are mounted on the rotor surface, the distance between the stator slots and the flux-carrying material that is guaranteed as a result allows the use of solid material without an increase in additional losses. According to a likewise particularly preferred exemplary embodiment, the synchronous machine has a stator with a stator, the stator being designed to guide a primarily radial magnetic flux, in particular to avoid magnetic flux being guided in a tangential direction. It is therefore a so-called "yokeless" design of the stator, which in particular avoids a magnetic flux guidance in the circumferential direction. A magnetic yoke in the stator is not required, which reduces weight and iron losses. According to one embodiment, the stator of the stator has a radial yoke thickness which is less than 30%, preferably less than 20%, particularly preferably less than 10% of a total radial stator thickness. In a so-called "yokeless" design, a mechanical connection of the stator teeth is still provided in this way. which would not be necessary electromagnetically and via which no functionally relevant magnetic flux takes place. The term "yokeless" thus refers to the electromagnetic design of the stator. According to one embodiment, the winding is designed to be torsionally stiff in such a way that a torque acting on the stator core during operation of a radial flux double-rotor machine can be supported, in particular completely, via the torsionally stiff winding on the carrier element. In this way, all other types of force support devices, in particular for the stator core, can advantageously be omitted. According to one embodiment, the stator core is designed to guide a primarily radial magnetic flux. It is therefore a so-called "yokeless" design of the stator core, which in particular avoids a magnetic flux guidance in the circumferential or tangential direction. A magnetic yoke in the stator core is not required, which reduces weight and iron losses. According to one embodiment, the stator core has a radial yoke thickness that is less than 30%, preferably less than 20%, particularly preferably less than 10% of a total radial stator core thickness. In a so-called "yokeless" design, a mechanical connection of the stator teeth is nevertheless provided in this way, which is not necessary electromagnetically and via which no functionally relevant magnetic flux takes place. The term "yokeless" thus refers in particular to the electromagnetic flux guidance of the stator core. According to one embodiment, the winding is formed from conductor bars connected to one another, in particular in the manner of a truss. In particular, the conductor bars can be integrally bonded be connected, for example by welding or soldering. However, other connection techniques would also be conceivable. Two conductor bars are preferably connected at the ends of the conductor bars and all the conductor bars together form a framework. The framework formed with the conductor bars is advantageously designed to be torsionally rigid and designed for torque transmission around the central axis of the stator. Furthermore, the conductor bars are formed with a thickness sufficient for power transmission. In a wheel hub motor, the thickness of the conductor bars can be in the range of several millimeters, for example. In particular, it can be rods with a square profile with edge lengths of several millimeters. According to one embodiment, the winding has a radially inner layer of helically arranged conductor bars and a radially outer layer of oppositely helically arranged conductor bars. In this way, a framework is formed by the winding, which has a high torsional rigidity. The conductor bars of the inner layer and the conductor bars of the outer layer each describe a helical line whose winding directions or gradients are opposite to one another. A swept angle of the helical line between the beginning and end of a conductor bar in relation to the central axis of the stator is in particular designed in such a way that in a radial flux double-rotor machine, one conductor loop is formed per pole of the rotors. The swept angle to be provided can thus be calculated from the quotient of a complete revolution (2π or 360°) and twice the number of pole pairs ^. According to one embodiment, the radially inner layer and the radially outer layer of the winding each have the thickness of a single conductor bar. That is, a phase of wick- ment is formed with the cross section of a single conductor bar. Such a design of a winding according to the invention is made possible, among other things, by the special design of the radial flux double rotor machine, which suppresses the current displacement to the surface that is otherwise present in conductors by means of its magnetic symmetry. In this way, comparatively thick conductor cross sections are made possible and a relatively uniform current distribution over the cross section is nevertheless achieved. For example, the thickness of the conductor bars can be in the range of several millimeters. It can in particular be rods with a square profile with edge lengths of several millimeters, for example in the range from 2 mm to 6 mm, in particular in the range from 3 mm to 5 mm. Other cross-sectional shapes are also possible. According to one embodiment, the conductor bars are twisted in accordance with the helical course in such a way that a cross section of a conductor bar is the same at every point of the conductor in relation to a radial axis of the cross section. In particular, it is a matter of torsion of a conductor bar, in particular a non-round one, around the central axis of the stator or of the machine. Depending on the course of the helical shape, the conductor bars can also be bent. The inner and the outer layer are interlaced with one another, that is to say twisted, twisted and possibly bent in opposite directions. In this way, the alignment of a conductor bar from a mechanical point of view at every point of the stator core is ideally aligned for power transmission with the stator core, so that the respective conductor bar is loaded evenly over its length. In the resulting framework, the conductors, when subjected to a tangential force, advantageously primarily absorb tensile and compressive stresses. In this way, peak loads and deformation of the conductor bars are avoided. the. The mechanical stresses can thus be significantly reduced, in particular compared to a design with axis-parallel, straight conductors. According to one embodiment, the conductor bars of the radially inner and outer layer belonging to the same phase of the winding are connected to one another at the ends of the conductor bars, in particular via a radially arranged conductor bar piece and/or by means of an integral connection. In addition to a conductor loop, this also results in a torsionally rigid truss-like structure, so that when an axially accessible winding end is fixed, the winding can absorb a high torque without causing impermissibly large deformations and/or stress states. Thus, the self-supporting design of the winding is made possible only by the winding material, for example copper, without additional support means or elements. According to one embodiment, the stator core contains a stator core with stator slots running helically in accordance with the course of the winding, with a single conductor bar being arranged in each stator slot of the stator core. The winding or the self-supporting framework formed with it is thus embedded in the laminated core of the stator. Similar to the conductor bars of the winding, the stator slots therefore change their tangential position depending on the axial position, resulting in the helical shape. The direction of the change in position follows the conductor bars, i.e. the center line of the radially outer grooves and the radially inner grooves also each describe a helix whose winding directions are opposite. In further embodiments, other types of production known to the person skilled in the art would also be suitable for producing the invention. Stator core geometry according to the invention with the radially inner and outer stator slots running in opposite helical configurations is conceivable, in particular also additive production methods, such as sintering methods or the like. According to one embodiment, only a single conductor bar is placed in each stator slot of the stator core. As already explained in relation to the winding, the conductor bars of the inner and outer stator slots are helically twisted against one another by torsion around the central axis of the machine, so that the conductor ends of the inner and outer layers are guided towards one another. The conductor bars are conductively connected to one another at the conductor bar ends, in particular via a radially arranged conductor bar piece and/or by means of an integral connection, for example by welding or brazing. According to one embodiment, the conductively connected conductor bars of the inner and outer layers together form wave-shaped winding strands. The winding strands can be interconnected by appropriate connections known to those skilled in the art to form a rotating field-generating winding with a desired or adaptable number of strands. The voltage-carrying number of phase windings results directly from the quotient of the number of slots in the meter and a product of the number of phases and the number of parallel branches in the meter. Advantageously, the number of parallel branches is chosen to be 1. In this case, the simplest possible wiring of the winding results. According to one embodiment, the stator laminations of the stator laminations are each of identical design with recesses provided for forming the stator slots. The helical course of the stator slots is provided by means of stacking of the stator laminations that are twisted relative to one another. In this way, the laminated stator core can be manufactured in a very economical manner, since the same stamping die can be used for all the stator laminates arranged or stacked in parallel. Accordingly, two adjacent stator laminations are rotated slightly relative to one another by a predetermined angle about the central axis, so that the recesses are arranged in an overlap relative to one another, which corresponds to the course of the helical line. According to an advantageous embodiment, the laminated core of the stator contains an inner sub-package with radially inner stator slots and an outer sub-package with radially outer stator slots. The stator laminations of the inner subpackage each have the same geometry and the stator laminations of the outer subpackage each have the same geometry. The stator laminations of the inner sub-package and the stator laminations of the outer sub-package are stacked in opposite directions to one another. In this way, the opposite helices of the stator slots can be realized with little manufacturing effort. Nevertheless, a still very economical production method is made possible, since the same stamping die can be used for all parallel arranged or stacked stator laminations of the inner part package and the same stamping die can be used for all parallel arranged or stacked stator laminations of the outer part package can. Correspondingly, two adjacent stator laminations of the inner subpackage are rotated slightly to one another by a predetermined angle about the central axis in a first direction and two adjacent stator laminations of the outer subpackage are rotated slightly to one another by a predetermined angle about the central axis in a second opposite direction. In this way, the recesses in the stator laminations of the inner sub-package and the recesses in the stator laminations of the outer sub-package overlap in opposite directions to one another arranged, which corresponds to the opposite course of the helix. According to a further embodiment, the stator laminations with recesses provided for forming the stator slots are each designed differently. The helical course of the stator slots is provided by means of different distances between the recesses in the individual stator laminations. In this respect, for each position of a stator lamination within the stack, an individually fitting stator lamination shape is produced, it also being possible for the individual geometries to be repeated within the stack. In this case, the production can be implemented, for example, by means of a beam cutting process, in particular a laser beam cutting process, which is more flexible in terms of shape compared to a stamping process. Also conceivable would be flexible stamping dies with variable geometry or, in the case of very large quantities, of course, several individual stamping dies for each of the different stator sheet metal shapes. According to a further development, the recesses for radially inner and radially outer stator slots are each integrated into a common stator lamination, the opposite helical progression of the radially inner and radially outer stator slots being caused by a continuous displacement of the inner and outer stator slots relative to one another Stator lamination is provided to stator lamination. Here, too, an individually fitting stator sheet shape is produced for each position of a stator sheet within the stack, with the individual geometries also being able to be repeated within the stack. Here, too, flexible cutting processes, such as laser beam cutting, are used for production. Through the possible Partial production of the inner and outer recesses is advantageously reduced the number of parts. According to one embodiment, the stator laminations have straight edges, in particular stamped edges. A width of the recesses provided for the stator slots is designed to be greater than the width of the conductor bars by an amount predetermined by the gradient of the helical shape of the course of the stator slots and by the sheet thickness of the stator sheets. A reduced clear width or continuous width of the stator slots due to the offset between the recesses of the stator laminations thus essentially corresponds to the width of a conductor bar. In practice, the continuous clear width of the stator slot is intended to be slightly larger than the width of the conductor bar in order to provide a clearance fit necessary for inserting the conductor bars. The edge of a stator slot thus describes a stair shape with the respective sheet thickness as steps, on which the conductor bar is evenly supported. In this way, the torque support is made possible uniformly over the entire thickness of the laminated stator core or over the entire length of the conductor bars accommodated in the laminated stator core. According to one embodiment, an angle swept by the stator slots is smaller than an angle swept by the conductor bars. The angle drawn relates to a rotation around the central axis of the stator. The difference in the swept angles comes about because the conductor bars protrude axially beyond the stator core and are therefore longer than the stator slots. Since the helical course also continues, the result is a larger swept angle. The said difference is intended to ensure sufficient accessibility of the winding ends for connecting, in particular welding, the conductor bar which is guaranteed after insertion into the stator slots. Furthermore, in this way, an axially offset engagement of the winding with the carrier device or its carrier element with respect to the stator core is made possible. From the quotient of the swept angles, ie a ratio of the angle swept by the stator slots to the angle swept by the conductor bars, a so-called degree of pole coverage for the stator core can be defined. According to one embodiment, a ratio of the angle swept by the stator slots to the angle swept by the conductor bars is in a range between 0.6 and 0.8, in particular between 0.6 and 0.75, preferably between 0.6 and 0.7. In this area, this ratio (degree of pole coverage) provides an optimum between losses caused by Joule heat and torque utilization. According to one embodiment, the carrier device has a carrier element in which carrier grooves are provided which correspond to the helical arrangement of the conductor bars and which engage with the conductor bars. In this way, a form-fitting embedding of the conductor bars in the carrier element is provided for supporting the torque at the axial end. There is preferably an engagement with all the conductor bars, so that the torque support is carried out homogeneously or evenly over the entire bar structure of the winding. In order to transmit the torque, the carrier element can be coupled to a mechanically fixed base of a radial-flow double-rotor machine. A possible design provides through-holes for non-positive fastenings. means such as screws, but of course form-fitting connection means or a bonded connection would also be conceivable. According to one embodiment, the carrier grooves follow the helical course of the twisted conductor bars at least in sections. In particular, the carrier grooves have a course that is twisted in the same way as the conductor bars. For example, the support element is essentially ring-shaped and has recesses on the inner and/or outer circumference, which are aligned radially and correspond to the course of the conductor bars. According to one embodiment, the carrier device has a radially inner carrier element for engagement with the radially inner layer of the conductor bars and a radially outer carrier element for engagement with the radially outer layer of the conductor bars. In this embodiment, the support elements can be ring-shaped, with the inner support element on its outer circumference corresponding to the course of the inner layer of the conductor bars grooves or teeth for positively receiving the radially inner conductor bars and the outer support element on its inner circumference the course of the outer layer of the conductor bars has corresponding grooves or teeth for positively locking reception of the radially outer conductor bars. The grooves or teeth follow in particular the respective helical course. Due to the arrangement on the inner or outer circumference, the recessed grooves are easily accessible for mechanical processing, which simplifies the production of the carrier elements. According to one embodiment of a radial flux double rotor machine, the support elements are fixed to the base and thus guide the torque to the fixed part of the electric machine. The carrier elements can be attached individually to the base, for example a housing, of the machine. Alternatively or additionally, the inner and outer support elements can also be fastened to one another. According to one embodiment of a stator, the carrier device contains a heat-conducting material, in particular a metal, preferably an aluminum alloy. In particular, both carrier elements can contain such a material. In this way, in addition to high mechanical strength, it is also possible to dissipate heat from the winding via the carrier device. According to one embodiment of a corresponding radial flux double-rotor machine with a carrier device containing a heat-conducting material, the base also has a heat sink which is designed to absorb heat dissipated from the stator, in particular from the winding, via the carrier device. As a result, the carrier device has a high mechanical strength and at the same time ensures a good thermal connection of the winding to the heat sink. For example, the housing of the machine can serve as a heat sink. Alternatively or additionally, the carrier device, preferably the inner and outer carrier elements, can be in thermal contact with an actively cooled heat sink of the machine. In this way, the current heat losses occurring in the winding or in the conductor bars can be effectively dissipated. According to one embodiment of a radial flow twin rotor machine, a predetermined number of pool pairs are provided on both the first rotor and the second rotor. An angle swept out by the conductor bars is used to form a conductor loop per pole of the rotors formed. The swept angle to be provided can thus be calculated from the quotient of a complete revolution (2π or 360°) and twice the number of pole pairs ^. The above configurations and developments can be combined with one another as desired, insofar as this makes sense. Further possible refinements, developments and implementations of the invention also include combinations of features of the invention described above or below with regard to the exemplary embodiments that are not explicitly mentioned. In particular, the person skilled in the art will also add individual aspects as improvements or additions to the respective basic form of the present invention. CONTENTS OF THE DRAWING The present invention is explained in more detail below with reference to the exemplary embodiments given in the schematic figures of the drawings. 1 shows an electric drive system according to the invention on the basis of a block diagram; FIG. 2 shows an electric drive system according to one embodiment based on a block diagram; FIG. 3 shows an example of an electric machine of the electric drive system according to the invention according to FIG. 1 using a schematic cross-sectional illustration; Fig. 4 based on a block diagram, a three- or multi-stage inverter circuit for an inventive the electric drive system according to the invention according to FIG. 1; 5 shows a particularly preferred exemplary embodiment of an inverter circuit according to the invention using a circuit diagram; 6 uses a flowchart to show a method according to the invention for operating an electric drive system; 7 shows a schematic longitudinal sectional illustration of a stator; 8 shows a schematic longitudinal sectional view of a radial flow double rotor machine; 9 shows an exploded view of a stator according to an embodiment; 10 is an exploded view of a radial flux dual rotor machine according to one embodiment; 11 is an exploded view of a radial flux dual rotor machine according to another embodiment; FIG. 12 shows a perspective illustration of the radial flux double rotor machine according to FIG. 11 in the assembled state; 13 is a longitudinal cross-sectional perspective view of a radial flow twin rotor machine according to yet another embodiment; 14 shows an exploded view of a stator lamination stack of a stator core; 15 shows a schematic longitudinal sectional illustration of a stator slot; 16 is a perspective view of a winding; Fig. 17 is a plan view of a winding; 18 shows a perspective view of an FEM simulation of a winding under load; 19 shows a perspective representation of an FEM simulation of a comparison winding with a straight design of the conductor bars under load; and FIG. 20 shows a flow chart of a method for manufacturing a stator. The accompanying drawings are provided to provide a further understanding of embodiments of the invention. They illustrate embodiments and, in connection with the description, serve to explain the principles and concepts of the invention. Other embodiments and many of the stated advantages will become apparent in view of the drawings. The elements of the drawings are not necessarily shown to scale with respect to one another. In the figures of the drawing, elements, features and components that are the same, have the same function and have the same effect—unless otherwise stated—are each provided with the same reference symbols. DESCRIPTION OF EXEMPLARY EMBODIMENTS FIG. 1 uses a block diagram to show an electric drive system 10 according to the invention for a motor vehicle. The electric drive system identified here by reference numeral 10 is preferably—but not necessarily—provided for use in a motor vehicle. The drive system 10 comprises at least one polyphase electrical synchronous machine 11 and an inverter circuit 12. The synchronous machine 11 is shown symbolized in the block diagram with a section of a cross-sectional diagram. It is connected on the input side to the inverter circuit 12 which drives the machine 11 . The synchronous machine 11 is designed as a double rotor machine and accordingly has a double rotor with two rotors 21 , 22 . Furthermore, a stator with a stator core (2) and a distributed winding (3) placed in the stator core (2) is provided, which is designed to be self-supporting for torque support. The double rotor is or the rotors 21, 22 are made of flow-carrying material made of solid material. According to the invention, the inverter circuit 12 is designed as a three-stage or multi-stage inverter circuit 12 . The inverter circuit 12 has at least one inverter 13 . The inverter 13 is coupled to the electrical machine 11 via its load output 15 and to a supply voltage source 18 via supply connections 16 , 17 . In this case, the inverter 13 is designed to convert a direct voltage VDC received on the supply side into an alternating voltage VAC. The inverter 13 is designed as a multi-phase inverter 13, the number of phases of the inverter 13 typically corresponding to the number of phases of the electrical machine 11. The electrical machine 11 is driven via the phase currents provided by the inverter 13 at the load output 15 . The synchronous machine is preferably a yokeless double rotor machine. The torsionally stiff winding is designed as a distributed winding and has a correspondingly low upper field spectrum. Due to this design, the rotors can be made of solid or solid material, since the winding generates only small upper fields and the resulting eddy currents in the rotor. Induced current harmonics, which would cause eddy currents and additional losses in the rotors made of solid material with a conventional two-stage converter, can be reduced with the three-stage or multi-stage inverter circuit 12 . By operating the three-stage or multi-stage inverter circuit 12, the causative harmonics in the input voltage can be significantly reduced, which leads to a reduction in the losses by more than 75%. FIG. 2 uses a block diagram to show an electric drive system 10 according to an embodiment. In the embodiment shown here, the operating mode of the inverter circuit 12 is set via an operating mode setting device 14 which, on the input side, has, among other things, is coupled to the electrical machine 11, adjustable. In particular, the operating mode setting device 14 can be used to set whether the inverter 13 is operating in two-stage operation, in three-stage or multi-stage operation or in mixed operation. Mixed operation designates an operating mode in which the inverter is operated both in two-stage operation and in three-stage or multi-stage operation, as can occur, for example, when changing from one operating mode to the next. The construction and functioning of the operating mode setting device 14 is explained in detail below with reference to the following FIGS. 4 to 6. The electrical machine 11 is a synchronous machine 11, preferably, but not necessarily, a three-phase synchronous machine 11. In this case, the inverter circuit 12 preferably includes a three-phase inverter 13. It is also preferred if the electrical machine 11 of the electrical drive system 10 is a Wheel hub motor for an electrically operated motor vehicle. However, other applications would also be conceivable and advantageous. 3 shows an example of a synchronous machine 11 of the electric drive system according to the invention according to FIG is and that the double rotor is also constructed from flux-carrying material made of solid material. The cross section of the double rotor synchronous machine 11 is shown in FIG. The double rotor machine 20 comprises the outer rotor 21 and the inner rotor gate 22. Between the two rotors 21, 22, the stator 23 is arranged in a manner known per se. The stator 23 can preferably, but not necessarily, be a yokeless stator 23 . The outer rotor 21 and inner rotor 22 are preferably not laminated, but constructed from solid material. The inner rotor 22 is tubular. However, a massive, full-volume configuration of the inner rotor 22 would be conceivable. It would be conceivable and advantageous if the magnets 24, 25 were embedded in pocket-shaped recesses provided specifically for this purpose in the outer rotor 21. However, it would also be conceivable for the magnets 24, 25 to be spaced apart from the outer rotor 21, ie not attached directly to its inner surface. The flux lines 27 between the north and south poles of the opposite-pole magnets 24, 25 run here in the core material of the outer rotor 21. In the example shown, there are also two opposite-pole magnets 28, 29 between the inner rotor 22 and the stator 23 on the inner surface of the Inner rotor 22 placed in the inner air gap 30. Here, too, the magnets 28, 29 can be embedded in corresponding pockets of the inner rotor 22 or spaced apart from the inner rotor 22. The flux lines 31 between the north and south poles of the opposite-pole magnets 28, 29 run here in the core material of the inner rotor 22. The flux-carrying material in the outer and/or inner rotor 21, 22 preferably consists of solid iron or a corresponding solid iron alloy. 4 uses a block diagram to show a three-stage or multi-stage inverter circuit for an electrical drive system corresponding to FIG. 2. The inverter circuit 12 includes—as already explained with reference to FIG. or multi-stage inverter 13 and an operating mode setting device 14. A first supply potential V11, for example a positive supply potential, can be tapped at the first supply connection 16. A second supply potential V12, for example a negative supply potential or a reference potential, can be tapped off at the second supply connection 17. A DC supply voltage VDC=V11-V12 is therefore present between the supply terminals 16, 17. A multi-phase load current I1 can be tapped off at the load output 15, via which the various phases of the electrical machine 11 that can be connected via the load output 15 are operated. The controllable three-stage or multi-stage inverter 13 is arranged between the supply connections 16 , 17 and the load output 15 . The inverter 13 is designed to convert a direct current voltage VDC taken on the supply side into an alternating current voltage VAC in order to provide the multi-phase load current I1 at the load output. The inverter 13 has a first driver stage 40 and at least one second driver stage 41 . The second driver stage 41 is designed to output load currents to the To lead load output 15, which are smaller than the currents provided by the first driver stage 40 output load. The operating mode setting device 14 serves the purpose of setting and thus controlling the operating mode of the inverter 13 and thus of the entire inverter circuit 12 . In particular, the inverter 13 is designed to operate the inverter 13 either in a first operating mode in a three-stage or multi-stage operation or in a second operating mode in a two-stage operation. At least a third operating mode would also be conceivable, which includes a mixed form of two-stage operation and three-stage or multi-stage operation. The third operating mode would be conceivable and useful in particular in the case of a transition from the first operating mode to the second operating mode and vice versa. The operating mode setting device 14 controls the operating mode used for the inverter 13 depending on the overall efficiency of the entire electric drive system 10. The overall efficiency is a function of the detected phase current of the electric machine 11 and at least one other parameter influencing the overall efficiency and/or one further property of the electrical machine 11 that influences the overall efficiency. For the purpose of setting the respective operating mode used, the operating mode setting device 14 comprises at least one of the following devices: an evaluation device 42; - a first measuring device 43; - a second measuring device 44; - a control device 45. Evaluation device 42 is designed to optimize the overall efficiency of electric drive system 10 based on the phase current and the at least one additional parameter and/or the at least one additional property. This can be done in situ, for example, that is to say during the operation of the electric drive system 10 . However, the relatively computationally intensive calculation is preferably carried out in advance, for example by means of a suitable calculation (e.g. numerically or analytically) and/or using a predefined family of characteristics. For example, the numerical efficiency calculation for 2L operation and 3L operation as well as the mapping of the function with decision output is carried out in advance, i.e. offline. The choice of the better efficiency with the help of switching and the use of the lookup table to determine the efficiency can also—but not exclusively—be made more or less dynamically during operation. The evaluation device 42 has an optimization module 46 for the purpose of optimization. The optimization module 46 first calculates the overall efficiency. The overall efficiency is then optimized analytically or using a lookup table, for example using an optimization function, taking into account the phase current and the at least one further parameter and/or property. The operating mode setting device 14 also includes first and/or second measuring devices 43, 44. The first measuring device 43 has at least one sensor input 47, for example. The operating mode setting device 14 can be coupled to the electrical machine 11 via the sensor inputs 47 in order to record and record electrical or physical parameters of the electrical machine 11, such as the phase current, the temperature and/or the rotor speed of the electrical machine 11 to er- grasp. The second measuring device 44 is arranged in such a way as to detect the temperature and/or the intermediate circuit voltage of the inverter 13, for example. In addition, the supply voltage VDC can also be detected via the second measuring device 44 . The actual control of the inverter takes place via a control device 45 provided specifically for this purpose. The control device 45 sets the respective operating mode of the inverter 13, i.e. whether the inverter 13 is operated in three- or multi-stage operation or in two-stage operation. The control device 45 can, for example, control the inverter 13 in such a way that both driver stages 40, 41 are activated in three-stage or multi-stage operation and the second driver stage 40 is deactivated in two-stage operation. 5 uses a circuit diagram to show a particularly preferred exemplary embodiment of an inverter circuit according to the invention. The DC supply voltage VDC is present at the supply connections 16 , 17 , with the supply potential V11=VDC/2 being able to be tapped at the first supply connection 16 and the supply potential V12=−VDC/2 being able to be tapped at the second supply connection 17 . A configuration would also be conceivable in which a reference potential, for example the potential of the reference ground GND, is present at the second supply connection 17 . In that case, the supply potential V11=VDC would be tappable at the first supply connection 16 . An intermediate circuit 50 consisting of a series connection of two intermediate circuit capacitors 51, 52 is connected on the input side of the inverter 13. The intermediate circuit 50 acts as an energy store. The inverter 13 shown in FIG. 5 includes a T-type neutral point clamped inverter architecture. For this purpose, the first, outer driver stage in the illustrated case of a 3-phase inverter has three half-bridge circuits 53a-53c, which are also connected on the load side between the supply connections 16, 17 with regard to their load paths. The respective center taps 54a-54c of the half-bridge circuits 53a-53c each form an output load connection 15a-15c of the inverter 13. Each of the half-bridge circuits 53a-53c has a first controllable power switch T1, T2, T3, which is High-side switches are formed. These first power switches T1, T2, T3 are connected to the first supply connection 16. The first power switches T1, T2, T3 are designed to provide a first voltage level at the load output 15. Each of the half-bridge circuits 53a-53c also has a second controllable power switch T4, T5, T6, which are designed as low-side switches. These second circuit breakers T4, T5, T6 are connected to the second supply connection 17. The second power switches T4, T5, T6 are designed to provide a second voltage stage at the load output 15. The second, inner driver stage 41 is connected between the center tap 55 of the intermediate circuit and the output load connections 15a-15c—and thus the respective center taps 54a-54c of the half-bridge circuits 53a-53c. In the example shown, the second driver stage 41 each includes three circuit branches 56a-56c. Each of the circuit branches 56a-56c comprises a series connection of two controllable circuit breakers T7/T8; T9/T10; T11/T12, which are arranged antiparallel with respect to their load paths. The controllable circuit breakers T7/T8; T9/T10; T11/T12 are designed to provide a third voltage level, which is between the first and the second voltage level, at the load output 15a-15c. To control the respective controllable circuit breakers, the control device 45 has a first control unit 45a and a second control unit 45b. The first control unit 45a is designed to control the power switches T1-T6 of the first driver stage 40. The second control unit 45b is designed to control the power switches T7-T12 of the second driver stage 41. In the exemplary embodiment in FIG. 5, the inverter 13 has a hybrid design. In this case, the power switches of the inverter 13 are not manufactured using the same semiconductor technology and/or are not of the same switch type. In particular, in the example shown, the power switches T1-T6 are Si-IGBTs with Si freewheeling diodes. The power switches T7-T12 are in the form of SiC MOSFETs. Alternatively (not shown in FIG. 5), the power switches T7-T12 can be in the form of SiC MOSFETs and the power switches T1-T6 can be in the form of GaN MOSFETs. Alternatively (also not shown in FIG. 5), the power switches T7-T12 can be in the form of IGBTs with a freewheeling diode and the power switches T1-T6 can be in the form of GaN power switches, in particular GaN MOSFETs. Alternatively (likewise not shown in FIG. 5), in a so-called homogeneous inverter topology, all power switches T1-T12 of the inverter 13 can be of the same switch type and/or with the same semiconductor technology. gie be made, for example as GaN power switches, SiC power switches, such as SiC MOSFETs designed. FIG. 6 uses a flow chart to show a method according to the invention for operating an electric drive system. The electrical drive system, which can be a drive system according to FIG. 2, for example, has a synchronous machine equipped with a double rotor. The double rotor is made of flux-carrying solid material. In a first step S1, the overall efficiency of the electric drive system is determined, for example offline. For this purpose, the phase current of the electrical machine of the electrical drive system is first detected (S11). In addition, at least one further parameter (S12) influencing the overall efficiency and/or at least one further property (S13) influencing the overall efficiency of the electrical machine is determined. In a next step S2, the synchronous machine is operated from all this information. A controllable three- or multi-stage inverter circuit is used for this purpose. The controllable three- or multi-stage inverter of the inverter circuit is operated either in the three- or multi-stage operating mode S21 or in the two-stage operating mode S22, depending on the overall efficiency of the electric drive system and the parameters and properties influencing it. A mixed form of three-stage or multi-stage operation and two-stage operation would also be conceivable. Such a mixed mode of operation would be conceivable, for example, in the case of a transition from three-stage or multi-stage operation to two-stage operation advantageous, for example, to avoid hard switching. The latter could be associated with losses and thus reduced efficiency. 7 shows a schematic longitudinal sectional illustration of a stator 101. This is a basic sketch of a stator 101 for a synchronous machine 110 designed as a radial flux double rotor machine according to a further embodiment (see FIG. 8), in particular for a wheel hub motor. The stator has a stator core 102 , a winding 103 and a carrier device 105 . The stator core 102, the winding 103 and the carrier device 105 are rotationally symmetrical about the center axis M shown. The winding 103 is designed to be self-supporting to support the torque of the stator and protrudes beyond the stator core 102 at at least one axial end 104 . The carrier device 105 is arranged offset axially with respect to the stator core 102 and is positively connected to the winding 103 at at least one axial end 104 for torque support. In this way, a torque present at the stator core 102 during operation of a radial flux double-rotor machine 110 can be supported on the carrier device 105 by means of the self-supporting winding 103 . The winding 103 contains a low electrical resistance conductor material, preferably copper. The stator core 102 is preferably constructed from a soft-magnetic material for guiding the magnetic flux. The carrier device preferably contains a heat-conducting material, for example an aluminum alloy. Of course, the winding 103 is electrically isolated. 8 shows a schematic longitudinal sectional view of a radial flux double rotor machine. This is also a purely illustrative principle sketch. The synchronous machine 110 designed as a radial flux double rotor machine accordingly has a mechanically fixed base 111, a first rotor 112 and a second rotor 113 in addition to the stator 101 according to FIG. The stator core 102, the winding 103, the carrier device 105, the base 111, the first rotor 112 and the second rotor 113 are also constructed rotationally symmetrically about the central axis M shown. The winding 103 is designed to be self-supporting to support the torque of the stator 101 and projects beyond the stator core 102 at at least one axial end 104 and is supported on the base 111 via the carrier device 105 . For this purpose, the carrier device 105 is arranged offset axially with respect to the stator core 102 and is positively connected to the winding 103 at at least one axial end 104 for torque support. The carrier device 105 is in turn fastened to the base, so that the torque can be supported on the base 111 via the carrier device 105 . The first rotor 112 is arranged radially inside the stator core 102 and the second rotor 113 is arranged radially outside the stator core 102 . The base 111 can, for example, be designed as the housing of the machine and, purely by way of illustration, comprises an L-shaped structure which is shown with two legs 107, 108. The representation is not to be understood as conclusive, rather the base can have further components and/or structural sections. The first leg 107 runs essentially radially, the second leg 107 is essentially axial with the greatest distance from the center axis M. The carrier device 105 is shown running radially in one piece purely schematically, but it can also be provided in multiple pieces and/or with a different geometry designed for a form fit with the winding 103 be. The illustrated overlap of the winding 103 with the base 111 is purely due to the illustrative schematic representation and does not mean a direct connection. The winding 103 is preferably connected via the carrier element 105 to the base 111 for torque support. The radial flux double rotor machine 110 according to FIG. 8 can be used as a synchronous machine 11 in an electric drive system 10 according to one of FIGS. 1 or 2 and in connection with an inverter circuit 12 according to one of FIGS. 9 shows an exploded view of a stator 101 according to a further embodiment. The stator 101 has a winding 103, a stator core 102 and a carrier device 105, with an advantageous exemplary embodiment of these components being shown in more detail in perspective here. The winding 103 is made up of an inner and outer layer with a plurality of conductor bars 106 connected to one another in the manner of a framework. The conductor bars 106 in the inner and outer layers are arranged opposite one another in a helical manner and are cohesively coupled at the ends of the conductor bars to a radial conductor piece 117 connecting the inner and outer layers. The thickness of the inner and outer layer corresponds in each case to the thickness of a conductor bar 106. This means that the winding 103 is formed by a single conductor layer forming the conductor loop and having a comparatively large cross section in the form of a conductor bar 106 in each case. Due to the truss structure formed with the conductor bars, the winding is torsionally rigid and thus designed to be self-supporting for torque support. The conductor bars 106 accordingly form wavy winding strands and can be connected to form a rotary field-generating winding with any number of strands by means of appropriate connections known to those skilled in the art and therefore not described further, such as delta connection, star connection or the like. In the embodiment shown, the stator core 102 and the carrier device 105 are each constructed from two components, for example. For the assembly of the stator 101, the winding 103, the stator core 102 and the carrier device 105 are arranged in a nested manner. After assembly, the components are aligned coaxially with one another on the common central axis M. The two-part support device 105 here, for example, is arranged axially offset relative to the other components and forms the innermost and outermost component of the stator 101. It is an inner ring and an outer ring each formed with grooves for positive engagement with the conductor bars. The two-part stator core 102 shown here by way of example is formed with two stator lamination packets 118 twisted in a helical manner relative to one another, which will be discussed in more detail with reference to FIG. In further embodiments, the stator core 102 and the carrier device 105 can each also be designed in one piece or with more than two parts. 10 shows an exploded view of a radial flux dual rotor machine 110 according to one embodiment. The radial flux dual rotor machine 110 includes, in addition to the stator 101 components, a first rotor 112, second rotor 113, and a base 111. FIG. The first rotor 112 is arranged radially inside and the second rotor 112 is arranged radially outside of the stator core 102 . The rotors 112, 113 are preferably made of a soft-magnetic solid material and have permanent magnets, so-called surface magnets, as poles on the respective surface facing the stator core. In further embodiments, other rotors known to those skilled in the art can also be used, for example with buried magnets, squirrel-cage rotors or electrically excited rotors. The base 111 is only shown schematically here for the sake of clarity. As already described in the description of FIG. 8 , the base 111 is attached to the carrier device 105 in the assembled state. The base 111 is fixed mechanically in relation to a reference system, for example a carrier of a vehicle axle. The radial flux double rotor machine 110 according to FIG. 10 can be used as a synchronous machine 11 in an electric drive system 10 according to one of FIGS. 1 or 2 and in connection with an inverter circuit 12 according to one of FIGS. 4 to 6. 11 shows an exploded view of a radial flow twin rotor machine 110 according to a further embodiment. The radial flow twin rotor machine 110 here has essentially the same components as explained with reference to FIGS. On the left side of the figure, the stator core 102, the winding 103, the first rotor 112 and the second rotor 113 are shown in the assembled state. The carrier device 105 shown on the right is also made in two parts and differs in the design of the respective annular inner carrier element 127 and outer carrier element 128 . The carrier elements 127 , 128 are equipped with carrier grooves 126 here. These are provided on the inner circumference of the outer carrier element 128 and on the outer circumference of the inner carrier element 127 for engagement with the conductor bars 106 of the winding 103 . For this purpose, the carrier grooves 126 are designed to be angled axially in accordance with the helical course of the conductor bars or its pitch, so that they can be brought into engagement with the conductor bars 106 of the winding 103 . The carrier elements 127, 128 are preferably made from a conductive metal, particularly preferably from an aluminum alloy. The two-part design of the carrier elements 127, 128 makes it possible for the carrier grooves 126 to be easily accessible during manufacture for mechanical or machining processing. The inner support element 127 and the outer support element 128 are each provided with a plurality of bores 109 circumferentially for attachment to the base 111 . The bores 109 are here, for example, uniform along a pitch circle distributed around the circumference. The individual bores 109 are located somewhat outside of the main body of the support elements and the support elements 127, 128 therefore form a star shape on the circumference facing away from the winding. Of course, other distributions of the bores 109 as well as other types of fastening means for the connection to the base 111 are conceivable. FIG. 12 shows a perspective representation of a radial flow double rotor machine 110 according to FIG. 11 in the assembled state. The carrier device 105 is fastened via the bores 109, for example in a machine housing (not shown) as a base 111 and thus leads the torque to the mechanically fixed part of the radial flux double rotor machine 110. In this way, the radial flux double rotor machine 110 generated torque can be effectively supported. The attachment of the carrier device 105 is implemented using appropriate attachment means (not shown), for example screws. The conductor bars 106 of the winding 103 extend axially on both sides to the outside of the stator core 102 and the first and second rotors 112, 113. The helically arranged conductor bars 106 of the radially inner and outer layers are connected to one another outside the stator core 102, respectively. The carrier elements 127, 128 are shown engaged with the conductor bars 106 of the winding 103 here. It can be seen that a conductor bar 106 is placed in each carrier groove 126, so that all conductor bars are positively coupled to the carrier device. Thus, a supported via the winding 103 torque via the carrier device device 105 are supported on the base 111 attached to the bores 109. The radial flux double rotor machine 110 according to FIGS. 11 and 12 can also be used advantageously as a synchronous machine 11 in an electrical drive system 10 according to one of FIGS. 1 or 2 and in conjunction with an inverter circuit 12 according to one of FIGS. 13 shows a longitudinal cross-sectional perspective view of a radial flow twin rotor machine 110 according to yet another embodiment. This embodiment essentially corresponds to the assembly of a radial-flow double-rotor machine 110 according to FIG. 10, the components of which will be discussed in more detail below. The stator core 102 has an inner subpackage 123 and an outer subpackage 124 . The partial packages 123, 124 run annularly between the first and second rotors 112, 113. Due to the sectional view, it is also possible to see the inner and outer layers 114, 115 of the conductor bars 106 running within the partial packages 123, 124. The radial flux double rotor machine 110 shown is a so-called “yokeless” design in which the yoke between two teeth is not in the functionally relevant magnetic flux. Although a stator yoke 130 runs between the conductor bars 106, this only serves to mechanically hold the laminated stator core 118 together. A radial yoke thickness can be made correspondingly thin, which in the illustrated embodiment is, for example, about 10% of the total radial stator thickness. With the comparatively small yoke thickness, additionally undesired magnetic stray flux in the yoke is reduced. In further embodiments, the radial yoke thickness can be less than 30%, preferably less than 20%, particularly preferably less than 10% of the total radial stator thickness for this purpose. The carrier device 105 also has an inner carrier element 127 and an outer carrier element 128 here. The support elements 127, 128 are arranged here in an axially offset manner in relation to the stator 101 and the rotors 112, 113, as can be clearly seen. Furthermore, the positive engagement of the carrier elements 127, 128 with the conductor bars 106 of the inner and outer layers 114, 115 can be seen at least in sections. It is also easy to see here that the conductor bars 106 of the inner and outer layers 114, 115 are connected at the conductor bar ends 116 via a radially arranged conductor bar piece 117. The connection is preferably implemented as an integral connection, for example by laser beam welding. The surface magnets of the rotors 112, 113 can also be seen in section. The first rotor 112 has a plurality of permanent magnets mounted on its outer peripheral surface. The second rotor 113 has a plurality of permanent magnets mounted on its inner peripheral surface. A particularly advantageous embodiment results when the rotors are made of soft magnetic solid material and are made with surface-mounted permanent magnets. In this design, the rotors can be manufactured very cheaply and a high level of efficiency can be achieved. 14 shows an exploded view of the stator core 118 of the stator core 102. As already mentioned, the laminated stator core 118 of the stator core 102 has an inner sub-package 123 and an outer sub-package 124 . This serves to simplify production of the oppositely twisted stator slots 119 with the same inner and outer stator laminations 121, 122 which are stacked twisted relative to one another and are provided with recesses at the same points Stator laminations are provided with differently arranged recesses and are stacked in the order necessary to form the stator slots. In further embodiments, completely one-piece stator cores 102 are also conceivable, which can be produced additively, for example. In the illustrated two-part design, an inner diameter of the outer part package 124 is almost the same as the outer diameter of the inner part package 123. This makes it possible to arrange the inner part package 123 coaxially within the outer part package 124. The subpackages 123, 124 are made up of individual annular stator laminations 121, 122 stacked on top of one another. The stator laminations 121 of the outer partial assembly 124 are manufactured with recesses positioned distributed over the outer circumference for forming the outer stator slots 119 . The stator laminations 122 of the inner partial assembly 123 are manufactured with recesses positioned distributed over the inner circumference for forming the inner stator slots 120 . For example, manufacturing such stator laminations by stamping is advantageous because of the quality of the edges and the very low manufacturing costs. The inner and outer stator slots 119, 120 describe helical lines which run in opposite directions to one another with the same pitch and which are characterized by the drawn-in angle of the stator slots α. The swept angle of the stator slots α can be defined from the angle between the position of the same stator slot on one axial side of the stator core 102 and on the other axial side of the stator core 102 with respect to the central axis M . The stator slots 119, 120 are designed here, for example, as T-slots with a rectangular recess with a tapered opening. These are intended in particular for the form-fitting reception of conductor bars with a rectangular cross section. Of course, the geometry of the recesses or stator slots can be adapted to the conductor geometry. Other cross-sectional shapes would also be conceivable. Fig. 15 shows a schematic longitudinal sectional view of a stator slot 119, 120. The usable or continuous clear width a of the stator slots 119, 120 within the stator laminated core 118 is essentially the same as the width of the conductor bars 106 accommodated within the stator core 102 educated. The stator laminations 121, 122 have straight edges, in particular stamped edges. Due to the offset of the sheets relative to one another, a width b of the recesses provided for the stator slots 119, 120 is greater than the width d of the conductor bars 106 by an amount predetermined by the pitch δ of the helical shape of the course and the sheet thickness t . In Fig. 16, a conductor bar 106 is shown schematically with dashed lines in the stator slot 119, 120, the continuous clear width a of the stator slot 119, 120 being slightly larger than the width d of the conductor bar 106 and the width a of the recess to provide a loose fit in the stator lamination 121, 122 is in turn formed significantly larger than the clear width b. The sheet thickness t and the angle of attack δ of the slope of the groove course represent a noticeable influencing factor for the difference between the width b of the recess and the clear width a of the usable passage within the groove in the case of straight, for example stamped, sheet metal edges. The difference comes about , since the gradient angle on the one hand and the stair-like stepping of the laminated core on the other hand have to be compensated. A minimum size of the width a of the recess for the limiting case of infinitely thin metal sheets, that is, considering the pitch angle δ of the conductor bar, would be b=1/cos(δ)*d. In order on the one hand to additionally compensate for the sheet metal thickness that is actually present and on the other hand to provide a clearance fit which allows the conductor bars to be inserted, the width b of the recess is actually provided to be even larger. The width b of the recesses according to FIG. 15 is dimensioned such that a clear width a of the stator slots 119, 120 reduced by the offset between the recesses of the stator laminations has a predetermined loose fit with the width d of a conductor bar to be inserted into the stator slot 106 forms, but the contact is still close enough to serve for evenly distributed power transmission or torque support between the stator laminated core and the winding. Such a dimensioning is made possible, among other things, by the fact that on the one hand each stator lamination is designed with the same high edge quality and twisted with the same offset, and on the other hand only a single conductor bar 106 is placed in each stator slot 119, 120, the dimensions of which are constant. In particular, in the illustrated embodiment, the conductor bar 106 is a rectangular bar with an edge length or width of several millimeters, for example in the range from 2 mm to 6 mm, in particular in the range from 3 mm to 5 mm. It can preferably be a rectangular profile of 5 mm×3 mm. 16 shows a perspective view of a winding 103. The winding 103 is made up of said conductor bars 106, which run helically along the central axis M. For this purpose, the conductor bars 106 are not only arranged correspondingly crossed, but also twisted according to the course of the helical line. The swept angle β of the conductor bars 106 identifies the angle between the beginning and end of a conductor bar 106 relative to the central axis M. Since the pitch of the helix of the conductor bars 106 is equal to the pitch of the helix of the stator slots 119, 120, but the conductor bars 106 are longer than the stator slots are formed, a ratio of the respective swept angles α and β can be formed to characterize the geometric conditions, which is also referred to as the degree of pole coverage. to a To provide an optimum between magnetic losses and torque utilization of a radial flux double rotor machine, this ratio (degree of pole coverage) is preferably in a range between 0.6 and 0.75. The opposite twisting and torsion of the inner and outer radial layers 114, 115 of conductor bars 106 can also be seen here. The torsion is designed in such a way that the cross section in relation to a radial line through the center of the conductor bar is always the same at every point of the conductor bar, which is also referred to as 2.5 D geometry. Thus, the conductor bar ends of the inner and outer layers 114, 115 are superimposed in a like alignment. The conductor bars 106 of the radial inner and outer layers 114, 115 can thus be conductively connected in a simple manner, here for example via a radially running conductor bar piece 117, which is welded to the conductor bar 106. It should be noted that the winding shown here is not produced individually, but always in combination with the stator core 102, which will be discussed in more detail with reference to FIG. 17 shows a plan view of a winding 103. The exact radial alignment of the conductor bars at every point of their helical course, which is aligned in the region of the central axis M in the perspective shown, can be clearly seen in this view. The conductor bar ends 116 each form the connection point between the inner and outer radial layers 114, 115. In the embodiment shown, the winding has a total of twelve connection contacts 31, for example. At three-phase connection, a three-phase operation is preferably provided. However, the winding can be adapted in a manner known to those skilled in the art to other interconnections to form a rotating field-generating winding with any number of strands. FIG. 18 shows a perspective view of an FEM simulation of a winding 103 under load. This is essentially the winding geometry shown in FIG. 16, with minor simplifications for simulation purposes. The scale shown relates to the voltages within the winding, where, for example, in the case of a rectangular profile of the conductor bars 106 of 5 mm×3 mm, it can be a scale from 0 MPa to 30 MPa. In this example, the conductor bar ends are defined by a swept angle of the conductor bars β>0, that is to say they are arranged and formed in a helical shape or formed in a correspondingly twisted manner. At the axial end, at which the carrier device engages, a maximum torque of the correspondingly dimensioned radial flux double rotor machine 110, shown with a bold arrow, is plotted than 1000 Nm, in particular more than 1500 Nm, in a specific example about 2000 Nm, in another specific example about 5000 Nm. The stresses within the winding are distributed very homogeneously due to the helix geometry. Hardly any deformation can be seen despite the strong superelevation that has been set. Because of this execution are thus Stress peaks and thus deformation are significantly reduced. Due to the truss-like structure, when an axially accessible winding end is fixed, a high torque can be absorbed by the winding 103 in a self-supporting manner, without causing impermissibly large deformations and/or states of stress. This can be explained in particular by the fact that the conductor bars 106 in the framework absorb primarily tensile and compressive stresses when they are subjected to a tangential force. Compared to designs with axis-parallel, straight conductors, the mechanical stresses can be reduced significantly. FIG. 19 shows a perspective view of a comparison model with a straight design and with the conductor bars 106 running axially under load. Compared to Fig. 18, due to the straight design and the axial course of the conductor bars, there is a stress curve concentrated on the side shown on the left in Fig. 18 and a strong deformation of the conductor bars resulting from the locally high stress with large deflection at the side shown in Fig. 19 visible on the page shown on the right. The same stress scale and the same increase in deformation are set here as in FIG. 18, which shows the effect of the different structural arrangements on the torsional stiffness. 20 shows a flow chart of a method for producing a stator 1. The method comprises a first step of providing S1 a stator core 102 with radially outer stator slots 119 each describing a helical line and radially inner stator slots 120 each describing a helical line with opposite winding direction individual conductor bars 106 following the helical lines through the inner and outer stator slots 119, 120. The conductor bars are inserted in particular in the axial direction. Furthermore, a step of connecting S3 of the conductor bars 106 introduced into the inner and outer stator slots at the conductor bar ends 116 to form conductor loops is provided. Although the present invention has been fully described above on the basis of preferred exemplary embodiments, it is not limited to them, but can be modified in a variety of ways.

List of reference symbols stator core winding electric drive system electric machine, synchronous machine (three or more stage) inverter circuit (three or more stage) inverter operating mode setting device load output a-15c output load connection , 17 supply connections supply voltage source double rotor, double rotor machine outer rotor inner rotor stator , 25 opposite-pole magnets ( of the outer rotor) (outer) air gap (outer) flux lines, 29 magnets of opposite polarity (of the inner rotor) (inner) air gap (inner) flux lines first (outer) driver stage second (inner) driver stage evaluation device first measuring devices second measuring devices control device optimization module sensor input 50 intermediate circuit circuit 51, 52 intermediate circuit capacitors 53a-53c half-bridge circuit 54a-54c center taps 55 center tap I1 (multi-phase) load current S1, S2 method steps S11 - S13 sub-steps S21, S22 sub-steps T1 - T3 first power switch of the half-bridge circuit, high-side switch T4 - T6 second power switch of the half-bridge circuit, low-side switch T7 - T12 power switch VAC (output side) AC voltage VDC (input side) DC voltage V11 (positive) supply potential V12 (negative) supply potential, reference potential 101 stator 102 stator core 103 winding 104 axial end 105 support means 106 conductor bar 107 first legs 108 second legs 109 bore 110 radial flux double rotor machine 111 base 112 first rotor 113 second rotor 114 radially outer layer 115 radially inner layer 116 conductor bar ends 117 conductor bar piece 118 stator core 119,120 stator slots 121,122 stator cores 123 inner subpackage 124 outer subpackage 125 carrier element 126 carrier grooves 127 inner carrier element 128 outer carrier element 129 permanent magnet α swept angle stator grooves β swept angle ladder bars δ pitch a clear width b Width of the recess d width of a conductor bar M central axis t sheet thickness

Claims

1. Electrical drive system (10) for or in a motor vehicle, with: at least one synchronous machine (11; 110) which has a double rotor (21, 22; 112, 113) and one in a stator core (2; 102) placed distributed winding (3; 103), wherein the double rotor (21, 22; 112, 113) is constructed from flux-carrying material made of solid material, wherein the winding (3; 103) is designed to be self-supporting for torque support; and at least one three- or multi-stage inverter circuit (12), which is coupled to the synchronous machine (11; 110) at a load output and which is designed to convert a DC voltage taken up on the supply side into an AC voltage, via which via the Load output, the synchronous machine (11; 110) can be driven, the inverter circuit (12) having a controllable three- or multi-stage inverter (13). 2. Drive system according to claim 1, characterized in that the synchronous machine (11; 110) is a wheel hub motor for an electrically operable motor vehicle. 3. Drive system according to claim 1 or 2, characterized in that the synchronous machine (11; 110) is designed as a radial flux double rotor machine. 4. Drive system according to one of the preceding claims, characterized in that that the winding (103) projects beyond the stator core (102) at at least one axial end (104) and a carrier device (105) is provided which is offset axially with respect to the stator core (102) and is intended for positive engagement with the winding (103). the at least one axial end is designed for torque support. 5. Drive system according to Claim 4, characterized in that the synchronous machine (110) has a mechanically fixed base (111), the carrier device (105) for torque support being in positive engagement with the at least one axial end of the winding (103) and is supported on the base (111). 6. Drive system according to one of the preceding claims, characterized in that the double rotor (21, 22; 112, 113) has a radially inside the stator core (2; 102) arranged first rotor (22; 112) made of solid material and a radially outside of Stator core (102) arranged second rotor (21; 113) made of solid material. 7. Drive system according to one of the preceding claims, characterized in that the flux-carrying material in the rotor consists of iron or an iron alloy. 8. Drive system according to one of the preceding claims, characterized in that the inverter circuit (12) has an operating mode setting device (14) which is designed to switch the inverter (13) depending on an overall efficiency of the electric drive system from a three-o - the multi-stage operation into a two-stage operation and vice versa, the overall efficiency being a function of the detected phase current of the synchronous machine and at least one other parameter influencing the overall efficiency and/or another property of the synchronous machine influencing the overall efficiency. 9. Drive system according to one of the preceding claims, characterized in that the inverter circuit (12) has an operating mode setting device (14) which is designed to switch the inverter (13) as a function of an overall efficiency of the electrical drive system from a three or - convert the multi-stage operation into a two-stage operation and vice versa, with the overall efficiency being a sole function of the detected phase current of the synchronous machine or a function of at least one other property of the synchronous machine influencing the overall efficiency. 10. Drive system according to one of the preceding claims, characterized in that the inverter (13) has a first driver stage and at least one second driver stage, the second driver stage being designed to carry output load currents to the load output which are smaller than those of output load currents provided by the first driver stage. 11. Drive system according to claim 9 or 10, characterized in that the operating mode setting device (14) has a control device which is designed to control the inverter in such a way that in three- or multi-stage operation the first driver stage and the second driver stage are activated and in two-stage operation at least one of the driver stages is deactivated. 12. Drive system according to one of claims 10 or 11, characterized in that the first driver stage has at least one bridge circuit, in particular a half-bridge circuit, whose center tap forms the output load connection of the inverter circuit, each bridge circuit having at least one first power switch, which is connected to a first supply terminal and which is designed to provide a first voltage level at the load output, and each bridge circuit also has at least one second power switch which is connected to a second supply terminal and which is designed to at the load output to provide a second voltage level. 13. Drive system according to one of claims 10 to 12, characterized in that the second driver stage has at least one third circuit breaker whose load paths are connected in series between an intermediate circuit and the center tap of the first driver circuit and which are designed to to provide a third voltage level, which lies between the first and the second voltage level, at the load output. 14. Drive system according to claim 9, characterized in that the operating mode setting device (14) has an evaluation device which is designed to optimize the overall efficiency based on the phase current or the at least one further property. 15. Drive system according to one of the preceding claims, characterized in that that the inverter (13) includes a T-type neutral-point clamped inverter architecture. 16. Drive system according to claim 3, characterized in that the winding (103) is designed torsionally rigid in such a way that a torque acting on the stator core (102) via the torsionally rigid winding (103) on the carrier device ( 105), in particular completely, can be supported. 17. Drive system according to claim 16, characterized in that the stator core (102) is designed to guide a primarily radial len magnetic flux. 18. Drive system according to claim 17, characterized in that the stator core (102) has a radial yoke thickness which is less than 30%, preferably less than 20%, more preferably less than 10% of a total radial stator core thickness. 19. Drive system according to one of the preceding claims, characterized in that the winding (103) is formed from conductor bars (106) connected to one another, in particular in the manner of a truss structure. 20. Drive system according to claim 19, characterized in that the winding (103) has a radially inner layer (115) of helically arranged conductor bars (106) and a radially outer layer (114) of oppositely helically arranged conductor bars (106). 21. Drive system according to claim 20, characterized in that the radially inner layer and the radially outer layer of the winding (103) each have the thickness of a single conductor bar (106). 22. Drive system according to claim 20 or 21, characterized in that the conductor bars (106) are twisted according to the helical course in such a way that a cross section of a conductor bar relative to a radial axis of the cross section is the same at every point of the conductor (6). is. 23. Drive system according to one of claims 20 to 22, characterized in that the same phase of the winding (103) associated conductor bars (106) of the radially inner and outer layers are connected to each other at the conductor bar ends (116), in particular via a radially arranged conductor bar piece (117) and/or by means of an integral connection. 24. Drive system according to one of claims 20 to 23, characterized in that the stator core (102) contains a stator core (118) with the winding course corresponding to helical stator slots (119, 120), wherein in each stator slot (119, 120) of the stator core ( 118) a single conductor bar (106) is arranged. 25. Drive system according to claim 4 or 5 and one of claims 20 to 24, characterized in that the carrier device (105) has a carrier element (125) in which the helical arrangement of the Conductor bars (106) corresponding and with the conductor bars (106) in engagement support grooves (126) are provided. 26. Drive system according to claim 22 and 25, characterized in that the carrier grooves (126) at least partially follow the helical course of the twisted conductor bars (106), in particular have an equally twisted course.